Barclays: Implied Valuation Multiples for US AI Stocks Declined 17% on Average Within One Month
nashnova research
Barclays' August 4 report shows the implied AI valuation multiples of its 12-stock AI basket dropped 17% in a month, from 52× to 43×; the market is repricing AI's near-term delivery, not rejecting the theme itself.
How far have valuations fallen?
Since Barclays began tracking in February, the 12 AI stocks have fallen an average of 25% from their respective peaks. Nearly half now sit at the lowest implied AI P/E — the multiple isolating the "AI premium" inside the stock price — of the entire tracking period.
Over the past month alone, the basket's average implied AI multiple dropped another 17%, roughly 9 turns, from 52× to 43×.
This means → the market is not abandoning AI; it is saying last quarter's earnings were not strong enough to justify the run-up.
Why the pullback — are fundamentals deteriorating?
Barclays is explicit: the correction is not driven by weaker fundamentals. Stocks had rallied hard ahead of earnings, and actual results could not sustain further multiple expansion.
In plain terms = expectations ran too far ahead of delivery; when earnings could not close the gap, prices had to come back.
For context, the S&P 500 fell just 1% over the same window while the Nasdaq 100 fell 5% — the AI basket's drawdown is a sector-specific de-rating, not a broad market event.
Which names were hit hardest?
Optics and EMS — electronic manufacturing services, the companies that build hardware on behalf of brands — led the decline. From peak: Corning down 46%, Fabrinet 42%, Celestica 40%.
On implied AI multiples: Corning collapsed from 155× to 76×, Celestica from 75× to 64×, Flex from 36× to 27×.
Super Micro trades at an implied AI multiple of just ~2×. Barclays flags this as company-specific, not an AI-industry signal.
Yet fundamental forecasts are rising?
Barclays has revised its cloud-AI revenue CAGR estimate (2023–2027) for the basket from an initial ~53% to ~65%, above the prior cycle's ~62%.
Individual standouts: Dell's 2026 cloud-AI revenue growth forecast sits at 144%, Cisco at 127%, Corning at 90%.
This reflects a striking contradiction: long-term growth expectations are improving while near-term prices are falling. The market is not asking "will AI deliver eventually?" — it is asking "can you deliver this quarter?"
Where is the money rotating?
Over the same window (July 6–31), Barclays' non-AI coverage rose by low-double-digit percentages on average, sharply outperforming the S&P 500.
Garmin gained 27%, Littelfuse 27%, FLIR-parent Teledyne 25%, Motorola Solutions 9%, Apple 4%.
This means → capital is flowing out of richly valued AI names and into companies that were "overlooked" in the first-half rally but have solid fundamentals — investors are seeking safer ground.
What to watch next?
AI stocks are still up roughly 60% year-to-date, well above the S&P 500's 9% — but that gain has narrowed from a prior peak of 70%–80%.
Even at cycle lows, implied AI multiples remain well above each company's core-business valuation, meaning the market still assigns a meaningful AI premium.
In plain terms = the AI premium is intact but compressing. Whether the next round of earnings can backstop that premium is the key test for whether the sector stabilizes.
市场有风险，内容仅供研究参考，不构成投资建议。